Prediction of this years Oscar Winners

Oscar for best picture
Argo (93%)
Oscar for best director
Steven Spielberg (81%)
Oscar for best actor
Daniel Day-Lewis (99.1%)
Oscar for best actress
Jennifer Lawrence (68.3%)

A group of researchers at Microsoft Research in New York City has created a data-driven model to predict the winners of this year's oscars. The model is an advancement of their previous approach to predict the winners of the election in 2012. Back then the model showed a very high accuracy (correctly predicting 50 of 51 jurisdictions). The current model's predictions are shown on the right.

The model is based on a similarity data-driven approach. Rothschild (the main creator of the model) says, it doesn't differ much from predictions in politics:

"I approach forecasting the Oscars the same way I approach forecasting anything, including politics. I look for the most efficient data, and I create statistically significant models without any regard for the outcomes in any particular year. All models are tested and calibrated on historical data, with great pains taken to ensure that the model is robust to "out of sample" outcomes, not just what has happend in the past. The models predict the future,  not just the past."

This, again, proves the point of not the data being the problem, but putting it in the right semantical context and composing it in a meaningful way. After the award ceremony it turned out that actually 19 of the 24 categories were correctly predicted by the data-driven model: an accuracy of nearly 80 percent.